Testicular Leydig cells in vitro secrete only inhibin α-subunits, whereas Leydig cell tumors can secrete bioactive inhibin

Molecular and Cellular Endocrinology, 83 (1992) 105- Il.5

0 1992 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/92/$05.00

105

MOLCEL 02677

Testicular Leydig cells in vitro secrete only inhibin cx-subunits, whereas Leydig cell tumors can secrete bioactive inhibin

J.P. de Winter a, M.A. Timmerman a, H.M.J. Vanderstichele b, LA. Klaij ‘, A.J. Grootenhuis ‘, F.F.G. Rommerts a and F.H. de Jong a

u Department of Endocrinology and Reproduction, Medical Faculty, Erasmus Ukersity Rotterdam, Rotterdam, Netherlands, and ’ Innogenetics, Ghent, Belgium

(Received 10 August 1991; accepted 2 October 1991)

Key words: Leydig cell; Leydig cell tumor; Inhibin; Inhibin a-subunit; Activin; Sertoli ceil

Summary

The secretion of inhibin and inhibin-related proteins by testicular Leydig cells was studied by estimation of inhibin immunoreactivity and bioactivity in spent media of preparations of immature and mature rat Leydig cells and of tumor Leydig cells. Immature and mature rat Leydig cells expressed inhibin a-subunit mRNA and secreted immunoreactive inhibin. The immunoreactive material did not contain inhibin bioactivity as measured by an in vitro rat pituitary bioassay system. Results of pulse labeling with [ ““SJmethionine followed by immunoprecipitation indicated that the inhibin-related pro-

teins secreted by the immature Leydig cell preparations are 26 kDa and 44 kDa molecules. Mature rat Leydig cells only secreted the 44 kDa inhibin-related protein.

Tumor Leydig cells (rat H540 and mouse MAlO) secreted immunoreactive and bioactive inhibit-r, which could be immunoneutralized by an antibody against inhibin. In the culture medium of some H540 tumor Leydig cells 26 kDa and 42 kDa inhibin-related proteins and 30 kDa inhibin were detected. In

culture medium of other H540 tumor Leydig cells, not secreting bioactive inhibin, only 26 kDa and 42 kDa inhibin-related proteins were found. No activin bioactivity was detected in culture media of immature rat Leydig cells, H540 and MA10 tumor Leydig cells.

It is concluded that normal Leydig cells secrete inhibin a-subunits, while Leydig cell tumors can also secrete bioactive inhibin. Neither normal Leydig cells nor Leydig cell tumors produce activin.

Introduction

lnhibin is a glycoprotein hormone composed of an a-subunit which is disulphide-linked to either of two &subunits, PA or PB. Inhibin

suppresses follicle-stimulating hormone (FSH) secretion by the pituitary gland (for reviews see de Jong, 1988; Ying, 1988; de Kretser and Robert- son, 1989). Dimers of inhibin /3 subunits (PAPA or /3A/?B) possess FSH-stimulating activity (Ling et al., 1986; Vale et al., 1986) and were named activins. Finally monomeric inhibin a-subunits (18

Correspondence to: J.P. de Winter, Department of En-

docrinology and Reproduction, Erasmus University Rotter-

dam, P.O. Box 1738, 3000 DR Rotterdam, Netherlands.

kDa, 26 kDa and 44 kDa) have also been described (Knight et al., 1989; Robertson et al., 1989; Sugino et al., 1989).

IOh

In the testis, bioactive and immunoreactive inhibin is produced by Sertoli cells (Grootenhuis et al., 1990). Release of immunoreactive, but not bioactive inhibin, by Sertoli cells in vitro is stimulated by FSH (Grootenhuis et al., 19901, whereas human chorionic gonadotrophin (hCG) does not affect the secretion of immunoreactive inhibin by Sertoli cells in vitro (Bicsak et al., 1987). How- ever, in vivo administration of hCG or Iuteinizing hormone (LH) increased levels of immunoreactive inhibin in plasma of rats (Sharpe et al., 1988; Drummond et al., 1989) and men (Mctachlan et al., 1988; Burger et al., 1990). These results sug- gest a role for Leydig cells in inhibin secretion.

Indeed results of immunocytochemical studies (Bergh and Cajander, 1990; Teerds et al., 1991) and in situ hybridization (Roberts et al., 1989) indicated the presence of inhibin-like material and mRNA for inhibin subunits in Leydig cells. Furthermore, Risbridger et al. (1989) detected immun~reactive and bioactive inhibin in culture medium of Leydig cells from adult rats, whereas Lee et al. (1989) reported that immature pig and rat Leydig cells synthesize activin. The aim of the present study was to characterize the inhibin-related proteins secreted by Leydig cells using radioimmunoassay, in vitro bioassay. Western blotting and immunoprecipitation of [ “Slmethionine- labeled proteins.

Materials and methods

Immature Leydig cells were isoIated from 21- 22-day-old Wistar rat testes as described by Rom- merts et al. (1985). Cells were attached to the surface of plastic 150 cm’ flasks (Costar, Cam- bridge, MA, USA) or 6-well plates (Costar) during 1 h in RPM1 (Seromed, Biochrom, Berlin, Germany) containing 0.1% bovine serum albumin (Sigma, St. Louis, MO, USA). After washing, the cell preparation contained approximately 30-50% Leydig cells and less than 1% Sertoli cells. The cells were cultured for 24 h in the same medium with or without ovine LH (oLH; 1 or 100 ng/ml culture medium, NIH S-20) or human FSH (hFSH; 25 mU or 2.50 mU/ml culture medium; Metrodin, Serono, Geneva, Switzerland). Media were collected and assayed for inhibin immunore-

activity and cells were Iysed in 1 N NaOH. Fur- ther batches of medium were pooled, concentrated and exchanged against 0.01 M phosphate- buffered saline (pH 7.01, using Centriprep 10 filters, which have a molecular weight cut-off at 10 kDa (Amicon, Danvers, MA, USA). Inhibin immunoreactivity and bioactivity were measured in the concentrated and exchanged media. Cells were used to isolate mRNA after an overnight culture period. In separate experiments cells from 6-well plates were labeled with [ ~~S]methionine after overnight culture in RPM1 containing 0.1% bovine serum albumin.

Mature Leydig cells were prepared from adult (at least 12 weeks old) Wistar rat testes using a Ficoll centrifugation purification step (Rommerts et al., 198.5). Cells were treated like immature Leydig cells, but were attached during 1 h in RPM1 containing 1% (v/v) fetal calf serum (Gibco, Paisley, Scotland, UK). These cell preparations contained approximately 90% Leydig ceils.

H540 rat tumor Leydig cells, grown subcuta- neously in rats were isolated as previously described to obtain a preparation of more than 99% H540 tumor cells (Rommerts et al., 198.5). The cells were treated like immature Leydig cells. Cells were stimulated with 1000 ng oLH/ml culture medium. Two types of responses were noted. Some H540 tumor Leydig cell preparations produced bioactive and immunoreactive inhibin (A), whereas other preparations only secreted immunoreactive inhibin (B). The mechanism that caused this difference is not clear.

MA10 mouse tumor Leydig ceils (100% pure, Ascoli, 1981), were cultured in plastic 75 cm’ flasks (Costar) or 6-well plates (Costar), containing RPM1 with 10% horse serum (Gibco) until confluency. Medium was then replaced by RPM1 containing 0.1% bovine serum albumin and the cells were treated like immature Leydig cells.

Sertoli cells were obtained from testes of 21- 22-day-old Wistar rats (Oonk et al., 1985) and cultured in 6-weli plates (Costar) as described by Grootenhuis et al. (1989). After 1 day of culture, germ cells were removed by replacing the culture medium with 10% Eagle’s minimal essential medium (MEM, Gibco) for 2 min as described by Toebosch et al. (1989). After the hypotonic shock treatment the medium was replaced twice with

107

MEM. The cells were cultured for another 2 days with or without 500 ng ovine FSH (oFSH, NIH S-16)/ml culture medium, followed by labeling with [35S]methionine. Alternatively, Sertoli cells were cultured for 24 h with or without 500 ng oFSH/ml culture medium and lysed in 1 N NaOH. Culture media were assayed for inhibin immunoreactivity.

RNA isolation and Northern blot analysis Total RNA was isolated from the different

Leydig cell preparations by an extraction with 3 M LiCl and 6 M urea, as described by Auffray and Rougeon (1980), followed by extraction with phenol and chloroform. Samples containing 40 pg RNA were separated by electrophoresis in a denaturing agarose gel (0.7% agarose; Sigma, St. Louis, MO, USA) containing ethidium bromide and blotted on GeneScreen membranes (NEN, Boston, MA, USA) by diffusion. Blots were hy- bridized with probes for the inhibin subunits and actin as described earlier (Klaij et al., 1990).

Inhibin radioimmunoassay Inhibin-like immunoreactivity present in the

culture media of the different cell preparations was measured using an antiserum (No. 1989) against purified 32 kDa bFF inhibin and iodi- nated 32 kDa bFF inhibin as described by Robertson et al. (1988). These materials were kindly provided by Dr. G. Bialy (NICHHD, Bethesda, MD, USA). Bovine follicular fluid with an arbitrary potency of 1 U/pg protein (Groo- tenhuis et al., 1989) was used as an inhibin standard.

Inhibin bioassay The amount of bioactive inhibin in the culture

medium was determined by an in vitro rat pituitary bioassay system, measuring the suppression of basal FSH release (Grootenhuis et al., 1989). Results of the FSH assay have been expressed in terms of NIADDK-rat FSH-RP2. Charcoal- treated bovine follicular fluid was used as inhibin standard. In this bioassay system activin stimu- lates FSH release as demonstrated using recombinant activin A (see below). Inhibin bioactivity was immunoneutralized with a polyclonal antiserum against partially purified bovine inhibin (van Dijk et al., 1986).

Activin bioassay Activin bioactivity was measured using the

‘animal cap bioassay’, in which culture media of different Leydig cell preparations were tested on their ability to induce mesodermal tissue in animal cap explants from Xenopus blastula (mod- ified method of Symes and Smith, 1987). Spawn- ing was induced in Xenopus laecis by injection of 300 IU human chorionic gonadotropin (Pregnyl, Organon, Oss, Netherlands) in the dorsal lymph sac of the females. The animals were left at room temperature. Eggs were laid about 10 h after the injection. Embryos were obtained by artificial fer- tilization in 5% De Boers solution (110 mM NaCI; 1.3 mM KCl; 0.44 mM CaCl,, pH 7.2) and al- lowed to develop at room temperature to the blastula stage. Stage 8 embryos were chemically dejellied using 2% cysteine-hydrochloride (Jans- sen, Antwerp, Belgium; pH 7.8-8.1) and transferred to Petri dishes coated with 1% Noble agar (Difco, Detroit, MI, USA) and containing 0.05% normal amphibian medium (NAM). A disc of animal pole ‘test tissue’ from the center of the pigmented animal hemisphere was dissected out using electrolytically sharpened tungsten needles. After excision, the animal cap was immediately transferred with the blastocoel facing upwards in 0.75% NAM, containing 0.1% bovine serum albumin (BSA) (fraction V, Sigma) and the test factor. The conditioned media from Leydig cells were tested both before and after heat treatment (10 min at 95°C). Recombinant activin A (Huylebroek et al., 1990) was used as positive control. Both experimental and control caps were cultured at 22°C for 3 days. Six to ten explants were tested for each sample in each of four independent experiments.

Labeling of secreted proteins with [35S]methionine Leydig cell and Sertoli cell preparations in

6-well plates were incubated for 30 min in methionine-free RPM1 or MEM, respectively. Sub- sequently the media were replaced by 600 ~1 methionine-free RPM1 or MEM per well containing 45 PCi [“‘Slmethionine (Amersham, Bucking- hamshire, UK) per ml. Culture was performed with or without 100 ng oLH/ml (1000 ng oLH/ml in case of H540 tumor Leydig cells) or 500 ng oFSH/ml. After a 6 h labeling period media

I ox

were collected and centrifuged at 8000 x g for 5 min. Supernatants were used for immunoprecipitation and cells were lysed in 1 N NaOH. Inhibin-like proteins were immunoprecipitated by

addition of one of two different rabbit polyclonal

antisera against synthetic peptides. One (anti-ac- antiserum) was directed against the 22 N-termi-

nal amino acid residues of the a-subunit of 32

kDa bovine inhibin (Grootenhuis et al., 1989). The other (anti-pro-antiserum) was directed against the 20 C-terminal amino acid residues of

the pro-part of the rat inhibin a-subunit precursor. After overnight incubation with 5 ~1 of antiserum per ml culture medium at 4”C, 50 ~1

IgSorb (The Enzyme Center, Maiden, MA, USA)/ml culture medium was added. 30 min

later tubes were centrifuged for 20 min at 8000 x g. Pellets were washed 3 times with 1 ml 0.01 M phosphate-buffered saline (pH 7.0) containing 1 mM EDTA, 0.05% (w/v) sodium dodecyl sulfate (SDS) and 1% (v/v) Triton X-100 and twice with

0.001 M phosphate-buffered saline (pH 7.0). Pel- lets were taken up in 20 ~1 sample buffer, boiled

for 3 min, centrifuged at 8000 X g and loaded on SDS polyacrylamide gels. Immunoprecipitations were also performed with sera obtained from the same rabbits before immunization (pre-immune sera) to evaluate the specificity of the immunoprecipitation.

SDS-po&acrylamide gel electrophoresis and West- ern blotting

SDS-polyacrylamide gel electrophoresis (SDS- PAGE) of “S-labeled proteins was performed on 1.5 mm thick, S-15% gradient gels as described by Laemmli (1970) with or without prior reduc-

tion (1% (v/v) /3-mercaptoethanol) of the samples. In parallel lanes the Rainbow ‘“C-labeled molecular weight markers (Amersham) were separated for molecular weight determination. Gels were fixed in water/acetic acid/methanol (3 : 2 : 5, v/v/v), stained with Coomassie brilliant blue, destained in water/acetic acid/methanol (6.3 : 0.7 : 3, v/v/v) and dried on a BioRad (Richmond, CA, USA) gel dryer. Subsequently, Hyperfilm MP (Amersham) was exposed to the dried gel at -8o”C, using an intensifying screen.

Samples used for Western blotting and molecular weight markers (Pharmacia, Uppsala, Swe-

den) were separated on 0.75 mm thick, 15% gels

with the Mini-Protean II system (BioRad) as described by Laemmli (1970). The same system was used to blot the separated proteins onto nitrocel- lulose in 1 h at 100 V. Western blotting was

performed as described by van Laar et al. (19891, using the antisera also used in the immunoprecipitations.

Estimation of DNA The DNA content of the cell lysates was deter-

mined by a fluorometric assay using 3,5-di- aminobenzoic acid dihydrochloride (DABA) as a fluorescent dye (Aldrich-Chemie, Steinheim,

Germany). Samples were neutralized with 1 N HCl. A 50 ~1 portion of the neutralized sample was mixed with 50 ~1 1.5 M DABA and incu-

bated for 45 min at 60°C. Subsequently 2 ml 1 N HCl was added and the fluorescence of the samples was measured using a Perkin-Elmer fluo-

rimeter at wavelengths of 415 nm (excitation) and 500 nm (emission). Calf thymus DNA was used as a standard.

Results

Expression of inhibin-subunit mRNA Inhibin a-subunit mRNA of 1.6 kb, as re-

ported previously by Klaij et al. (1990), was expressed in all Leydig cell preparations examined after overnight culture (Fig. 1). The amount of inhibin a-subunit mRNA found in the mature

Leydig cell preparation was very small (Fig. 1, lane 9). However, immediately after isolation of the cells more a-subunit mRNA was detected (Fig. 1, lane 4). Inhibin pB-subunit mRNAs (4.2 kb and 3.5 kb respectively) were only detected in total testis RNA and Sertoli cell RNA of 21-day- old rats and in RNA isolated from tubuli of mature rats, but not in RNA from the examined Leydig cell preparations (Fig. 1). Inhibin DA-subunit mRNA was not detected in any of the examined preparations (result not shown).

Inhibin immunoreactivity Culture medium derived from Leydig cell

preparations and Sertoli cells was assayed for inhibin immunoreactivity by radioimmunoassay (RIA). Results of one experiment have been sum-

PB

CY

act

Fig. 1 Northern blot analysis of the expression of inhibin

123456789

@B-subunit mRNA, inhibin o-subunit mRNA and actin mRNA in 21-day-old rat testes (11, and preparations of immature rat Sertoli cells (21, immature rat Leydig cells (31, mature rat Leydig cells immediately after isolation (4), H540 (A) tumor cells (5), H540 (B) tumor cells (6), MA10 tumor cells (71, mature rat tubuli (8) and mature rat Leydig cells after

overnight culture (91.

TABLE 1

IMMUNOREACTIVE INHIBIN IN CULTURE MEDIA OF LEYDIG AND SERTOLI CELL PREPARATIONS AFTER 24 h OF CULTURE (MEANS f SEMI

Values have not been corrected for the purity of the cell preparations.

Cell type and additions n Inhibin (U/Kg DNA/24 h)

Sertoli cells - FSH 6 0.48 k 0.03 + FSH (500 ng/ml) 6 1.83k0.15 **

Immature Leydig cells -LH 6 0.25 f 0.02 + LH (100 ng/mI) 6 0.41+ 0.03 **

Mature Leydig cells -LH 3 0.14kO.02 + LH (100 ng/ml) 3 0.24 k 0.01 **

H540 (A) tumor cell -LH 6 4.10+0.23 + LH (1000 ng/mlI 6 4.06 + 0.28

MA10 tumor cells -LH 6 0.64 f 0.05 + LH (100 ng/ml) 6 0.47 f 0.05 *

* Significantly different (P < 0.05) from appropriate control (Student’s t-test).

+* Significantly different (P < 0.01) from appropriate control (Student’s r-test).

109

TABLE 2

EFFECT OF DIFFERENT CONCENTRATIONS OF OVINE LH (oLH) AND HUMAN FSH (hFSH) ON SECRE- TION OF IMMUNOREACTIVE INHIBIN BY IMMA- TURE RAT LEYDIG CELL PREPARATIONS (MEANS f SEM, n = 3)

Addition Inhibin (U/ml culture medium)

Control 0.24 + 0.03 1 ng oLH/ml 0.52 f 0.06 * 100 ng oLH/ml 0.56 + 0.02 * 25 mU hFSH/ml 0.22 * 0.04 250 mU hFSH/ml 0.33+0.10

* Significantly different (P < 0.02) from control (Student’s t-test).

marized in Table 1; two similar experiments yielded essentially the same results. Immature and mature rat Leydig cell preparations and Ser- toli cells secreted immunoreactive inhibin. Secre- tion of this immunoreactivity was significantly stimulated by LH (Leydig cells) or FSH (Sertoli cells), suggesting that at least part of the inhibin- like material is secreted by Leydig and Sertoli cells. In a separate experiment human FSH (Metrodin) had no significant effect on secretion of immunoreactive inhibin by immature rat Ley- dig cells (Table 2). H540 and MA10 tumor Leydig cells secreted a relatively large amount of inhibin-like material, but LH did not stimulate its secretion. In MA10 cells LH even appeared to inhibit release of inhibin-like material.

Inhibin bioactivity Immature rat Leydig cells, Addition of culture

medium obtained from immature rat Leydig cell preparations caused a stimulation of FSH release in the rat pituitary cell culture (Fig. 2). The same medium did also inhibit basal LH release (data not shown). After exchange of the medium against phosphate-buffered saline, FSH-release-stimulating and LH-release-inhibiting activity disappeared from the culture medium. This is in contrast with the effects of culture medium containing recombinant activin A, which stimulated basal FSH release both before and after exchange (Fig. 21, without effect on basal LH release. In order to avoid these effects further experiments were per-

110

4

1 C ilc activin

10 100 1000 10 100 1000

microliter medium added

Fig. 2. Effect of culture medium from an immature rat Leydig

cell preparation (ilc), culture medium containing recombinant

activin A (activin) and control medium (c) on basal FSH

release by rat pituitary cells, before (open symbols) and after

(closed symbols) exchange of the culture medium.

formed with exchanged media. Results of one of these experiments are shown in Fig. 3; in five parallel experiments, similar data were obtained. Addition of an immunoneutralizing antiserum

Fig. 3. Effect of exchanged culture media from different

Leydig cell and Sertoli cell preparations on basal FSH release

by rat pituitary cells ((0) rat Sertoli ceil culture medium, (A) immature rat Leydig cell culture medium, (v) mature rat

Leydig cell culture medium, ( q ) H540 (AI tumor cell medium,

(+) H540 (B) tumor cell medium, (+) MA10 tumor cell

medium).

TABLE 3

EFFECT OF INHIBIN IMMUNONEUTRALIZATION ON

FSH RELEASE BY RAT PITUITARY CELLS, AFTER

ADDITION OF CULTURE MEDIA OBTAINED FROM

PREPARATIONS OF IMMATURE RAT LEYDIG CELLS

AND DIFFERENT TUMOR LEYDIG CELLS (MEANS+

SEM, n = 3)

Addition

Control

Immature Leydig cell medium

FSH release (rig/ml)

19.9 * 2.2

without neutralization

with neutralization

H540 (A) tumor cell medium

20.5 + 4.0

21.0+ 1.7

without neutralization

with neutralization

MA10 tumor cell medium

6.X+_ 1.3

20.1 k3.8

without neutralization 4.7* 1.0

with neutralization 14.2kO.6

against inhibin did not affect FSH release in the presence of immature Leydig cell culture medium (Table 3) or culture medium containing recombinant activin A (not shown).

Mature Leydig cells. Exchanged culture medium from mature rat Leydig cell preparations, which contained immunoreactive inhibin, did not affect basal FSH release significantly. Results of one experiment have been included in

Fig. 3; in four parallel experiments also no specific suppression of FSH release was observed.

Tumor Leydig cells. Fig. 3 also shows the ef-

fect of addition of exchanged culture media, obtained from Sertoli cells and different tumor Ley- dig cell preparations, on basal FSH release by rat pituitary cells. Media from Sertoli cells, H540 (A) tumor Leydig cells and MA10 tumor Leydig cells suppressed basal FSH release. Suppression was parallel to that caused by addition of bovine follicular fluid. In contrast, the culture medium from the H540 (B) tumor cells did not suppress basal FSH release (Fig. 3). Addition of the immunoneutralizing inhibin antiserum abolished the effect of culture medium derived from H540 (A) tumor cells and MA10 tumor cells (Table 3).

Actizlin bioacticity. Analysis of conditioned media from immature rat Leydig cell preparations, H540 (A) and MA10 tumor Leydig cells in the animal cap bioassay did not show any meso- derm inducing capacity, indicating that less than

0.2 ng activin/ml culture medium was present (Smith et al., 1990). The same results were obtained after heating of the samples (data not shown).

Immunoprecipitation of [“SSlmethionine-labeled proteins

In order to visualize inhibin-like proteins, secreted by the various testicular cell types, proteins were labeled with [ 35S]methionine, immunoprecipitated and separated by SDS-PAGE (Fig. 4). With a polyclonal antiserum against the first 22 N-terminal amino acid residues of the cr-subunit of bovine 32 kDa inhibin, Sertoli cell secreted proteins of 44 kDa, 30 kDa and 26 kDa were specifically precipitated (Fig. 4a, lane B2). There were also proteins aspecifically precipitated (Fig. 4a, lane Bl), due to binding to IgSorb used to precipitate the antibodies (data not shown). The inhibin-like proteins of 44 kDa and 26 kDa were also specifically precipitated with a polyclonal antiserum against the 20 C-terminal amino acid residues of the pro-sequence of the rat inhibin a-subunit (Fig. 4c, lane B2). Their secretion was stimulated by FSH (Fig. 4a, lane B3). Similar 44 kDa and 26 kDa inhibin-related proteins were present after immunoprecipitation of [ 35S]methionine-labeled proteins secreted by an immature rat Leydig cell preparation (Fig. 4a, lane A2) and their secretion was stimulated by

Fig. 4. Immunoprecipitation of [35S]methionine-labeled pro-

teins secreted from Sertoli and Leydig cell preparations. For

description of the antisera see Materials and methods. Panel

a: Precipitation of proteins secreted by immature Leydig cells

(A). immature Sertoli cells (B) and H540 (A) tumor cells (C)

with pre-immune serum (lanes 1) or with the anti&c-antibody

(lanes 2 and 3). Cells were not stimulated (lanes 2) or stimu-

lated with ovine LH or ovine FSH (lanes 3). Panel b: Precipi-

tation of proteins secreted by mature Leydig cells with pre-im-

mune sera (lanes 1) or specific antisera (lanes 2). Precipitation

A represents proteins secreted without stimulation and pre-

cipitated with the ant&c-antibody. Precipitation B represents

proteins secreted after stimulation with LH and precipitation

with the same antibody. Precipitation C shows proteins se-

creted without stimulation after precipitation with the anti-

pro-antibody. Panel c: Precipitation of proteins secreted by

H540 (A) tumor cells (A) and immature Sertoli cells stimu-

lated with oFSH (B) with pre-immune serum (lanes 1) or the

anti-pro-antibody (lanes 2).

111

(a) A 8 C 123123123

(b)

69-

46-

30-

21-

112

LH (Fig. 4a, lane A3). Only the 44 kDa inhibin- like protein was detected in culture medium of a mature rat Leydig cell preparation after precipitation with both antisera (Fig. 4b, lanes A2 and C2); its secretion was stimulated by the addition of LH (Fig. 46, lane B2).

H540 (A) tumor Leydig cells secreted inhibin- like proteins of 42 kDa, 30 kDa and 26 kDa (Fig. 4a, lane C2) but in culture medium of H540 (B) tumor Leydig cells only the 42 kDa and 26 kDa proteins were detected (results not shown). This 26 kDa protein was also precipitated with the antiserum against the 20 C-terminal amino acid residues of the pro-sequence (Fig. 4c, lane A2). Addition of LH did not affect the secretion of these proteins (Fig. 4a, lane C3).

After reduction of the samples the 26 kDa protein disappeared, while a 18 kDa protein was detected (results not shown).

Western blotting Western blotting was not sensitive enough to

detect inhibin-like proteins in culture media of immature and mature rat Leydig cell preparations. However, this technique showed the 26 kDa inhibin-related protein in culture media of H540 (A) tumor Leydig cells and Sertoli cells with both antisera used (Fig. 5). This 26 kDa protein was reduced with 1% /?-mercaptoethanol to give an 18 kDa protein, which was immunoreactive with the antiserum against the first 22 N-terminal amino acid residues of the a-subunit of bovine 32 kDa inhibin (result not shown). Also 30 kDa inhibin was detected in culture medium

Fig. 5. Western blot of Sertoli cell culture medium (1) and

H540 (A) tumor Leydig cell medium (2) with an antibody

against the LYC part (A) or against the pro-part (B) of the

inhibin a-subunit.

of both cell types (Fig. 5) with the antiserum against the ac-part of the inhibin molecule.

Discussion

There is growing evidence that production of inhibin and inhibin-related proteins is not re- stricted to Sertoli cells and granulosa cells. Meu- nier et al. (1988a) showed expression of inhibin a, PA and PB subunits in various extragonadal tissues. These authors also detected inhibin LY- subunit mRNA in different rat ovarian cell types, including interstitial cells (Meunier et al., 1988b). The results of the present study indicate that interstitial cells of immature and mature rat testes also can produce inhibin subunits, in agreement with the observations of Roberts et al. (1989). In immature rat Leydig cell preparations expression of inhibin a-subunit mRNA was detected, result- ing in inhibin immunoreactivity in culture medium of these cells. The measured immunoreactive material is not likely to be produced by contaminat- ing Sertoli cells, because human FSH did not affect secretion of immunoreactive material in this Leydig cell preparation (Table 2).

The immunoreactive material did not cause FSH suppression in the in vitro bioassay for inhibin. The results of the immunoneutralization experiment (Table 3) argue against the possibility that similar amounts of bioactive inhibin and activin are produced by immature Leydig cells, and contrast with the data of Lee et al. (1989), who detected FSH-release-stimulating activity in the culture medium of immature pig and rat Leydig cells. However, these authors did not exchange the culture medium of the Leydig cells, in order to exclude effects of small molecular weight substances present in these media. In the present study the FSH-release-stimulating activity found before exchange disappeared after exchange of the media, whereas the activity of recombinant activin A was not’affected by this exchange. Fur- thermore the results of the animal cap bioassay and the absence of inhibin P-subunit mRNAs make it unlikely that immature rat Leydig cells produce activin.

The immunoreactive material present in immature rat Leydig cell culture medium corre-

113

sponds to inhibin-related proteins of 44 kDa (presumably procrNaC) and 26 kDa (presumably pro&), which are both encoded by the inhibin a-subunit precursor mRNA. Those proteins do not possess inhibin bioactivity as shown earlier by Grootenhuis et al. (1990) for the 26 kDa protein. Furthermore, in culture medium of H540 (B) tumor Leydig cells, which only secreted similar 42 kDa and 26 kDa inhibin-related proteins, no inhibin bioactivity was detected.

Mature rat Leydig cells express inhibin a-subunit mRNA and immunoreactive inhibin is measured in culture medium of these cells. This immunoreactive material is devoid of inhibin bioactivity since specific suppression of FSH release was not found after addition of this medium to cultured pituitary cells. The absence of PA and PB mRNA expression correlates with this obser- vation. However, Risbridger et al. (1989) detected bioactive inhibin in mature rat Lcydig cell culture medium, and more recently described the presence of immunoreactive follistatin in mature Ley- dig cell culture medium (Risbridger et al., 1991). The reason for the differences between the present observations and those of Risbridger et al. (1989, 1991) are not clear.

Precipitation of [ 35S]methionine-labeled proteins revealed the immunoreactive protein, secreted by mature Leydig cell preparations to be a 44 kDa inhibin-related protein (presumably proaN&). It seems that Leydig cells lose their ability to process the inhibin a-subunit precursor with increasing age, because no 26 kDa inhibin- related protein was detected in mature rat Leydig cell culture medium, different from the situation in the culture medium immature rat Leydig cell preparations.

In contrast to normal Leydig cells, two out of three tumor Leydig cells did produce bioactive inhibin. These cells expressed inhibin a-subunit mRNA but no inhibin P-subunit mRNAs were detected; the expression of inhibin P-subunit mRNAs was probably below the detection limit of the Northern blot technique. The presence of bioactive inhibin, which could be immunoneutralized by an inhibin antiserum, indicates that inhibin p-subunit must be present. Expression of inhibin (Y- and /3B-subunit mRNAs in cultured MA10 cells, without expression of PA-subunit

mRNA was reported earlier by Feng et al. (1990). H540 tumor Leydig cells (A) secreted 30 kDa inhibin and 42 kDa and 26 kDa inhibin-related peptides. In culture medium of H540 (B) 30 kDa inhibin was not present. LH had no effect on the release of inhibin-related material from H540 tumor Leydig cells, perhaps because the level of LH receptors is only about 1% of that present in normal Leydig cells from adult rats (Erichsen et al., 1984). In these cells steroid production can also hardly be stimulated by LH. Finally, no activin bioactivity was detected in the medium of H540 and MA10 tumor Leydig cells. This con- trasts with the results of Lee et al. (1989) who observed FSH-stimulating activity in the medium of the immortalized mouse Leydig cell line TM3. However, using the animal cap assay, which has a higher specificity for activin bioactivity, meso- derm inducing activity could not be detected in culture media from TM3, R2C or I-10 Leydig cell lines (H.M.J. Vanderstichele, non-published results).

Inhibin-like material was detected by immuno- cytochemistry in human (Bergh and Cajander, 1990) and rat (Teerds et al., 1991) normal and tumor Leydig cells. Furthermore, Roberts et al. (1989) detected positive staining in normal rat Leydig cells and de Jong et al. (1990) found inhibin immunoreactivity in homogenates of human Leydig cell tumors. In contrast several authors (Merchenthaler et al., 1987; Rivier et al., 1988; Saito et al., 1989; Shaha et al., 1989) did not detect immunocytochemical staining for inhibin subunits in normal Leydig cells, probably because of the relatively low amounts of inhibin- like material produced by these cells.

Several reports have demonstrated that in vivo administration of hCG or LH increases serum immunoreactive inhibin levels in adult rats (Sharpe et al., 1988; Drummond et al., 1989) and men (McLachlan et al., 1988). In the adult rat this seems to be the result of a change in the polarity of inhibin-c-u secretion by Sertoli cells (Maddocks and Sharpe, 1990), rather than a contribution of Leydig cells. Destruction of Leydig cells by ethylene-1,2_dimethanesulfonate (EDS) did not lead to any decrease in the levels of immunoactive inhibin in blood and testis (Mad- docks and Sharpe, 1989). The possible role of

114

inhibin-related proteins produced by Leydig cells needs to be determined by further studies.

Nothing is known about the contribution of inhibin a-subunits from immature rat Leydig cells to serum immunoreactive inhibin levels. Part of the serum inhibin immunoreactivity in immature rats could be of Leydig cell origin.

We conclude that immature and mature Ley- dig cells and Leydig cell tumors can secrete inhibin-like material. In contrast to Sertoli cells and Leydig cell tumors, immature and mature Leydig cells do not secrete inhibin bioactivity, but only a-subunits. Immature Leydig cells secrete inhibin-like proteins of 44 kDa (presumably procuNcrC) and 26 kDa (presumably pro&), while mature Leydig cells only secrete the former inhibin-related protein. Neither normal Leydig cells nor Leydig cell tumors secrete activin.

Acknowledgements

This study was supported in part by the Dutch Organization for Scientific Research (NW01 through SB-MW (Medical Sciences). We wish to thank S. Pattijn for assistance in performing the animal cap bioassay and H.J. van Loenen for culturing the MA10 tumor Leydig cells, kindly given by M. Ascoli. We are grateful to the NIH for gifts of oLH, oFSH and to the NICHHD for the materials used in the inhibin radioimmunoassay. We would like to thank Dr. N. Ling and S. Shimasaki for their generous gift of the inhibin cDNA probes.

References

Ascoli, M. (1981) Endocrinology 108, X8-95. Auffray, C. and Rougeon, F. (1980) Eur. J. Biochem. 107,

303-314. Bergh, A. and Cajander, S. (1990) Int. J. Androl. 13, 463-469. Bisack, T.A., Vale, W., Vaughan, J., Tucker, E.M., Cappel, S.

and Hsueh, A.J.W. (1987) Mol. Cell. Endocrinol. 49, 211- 217.

Burger, H.G., Tiu, S.C., Bangah, M.L. and de Kretser, D.M. (1990) Reprod. Fertil. Dev. 2, 137-144.

de Jong, F.H. (1988) Physiol. Rev. 68, 555-607. de Jong, F.H., Grootenhuis, A.J., Steenbergen, J., van Sluijs,

F.J., Foekens, J.A., Ten Kate, F.J.W., Oosterhuis, J.W., Lamberts, S.W.J. and Klijn, J.G.M. (1990) J. Steroid B&hem. Mol. Biol. 37, 863-866.

de Kretser, D.M. and Robertson, D.M. (1989) Biol. Reprod. 40, 33-47.

Drummond, A.E., Risbridger, G.P. and de Kretser. D.M. (1989) Endocrinology 125, 510-515.

Erichsen, A.A., Jahnsen, T., Andersen, D., Torjesen, P. and Hansson. V. (1984) J. Steroid Biochem. 21,539-543.

Feng, Z.-M. and Chen, C.-L.C. (1990) The Endocrine Society 72nd Annual Meeting, Atlanta, The Endocrine Society, Bethesda, MD, abstract 770.

Grootenhuis, A.J., Steenbergen, J., Timmerman, M.A., Dors- man, A.N.R.D., Schaaper, W.M.M., Meloen, R.H. and de Jong, F.H. (1989) J. Endocrinol. 122, 293-301.

Grootenhuis, A.J., van Beurden. W.M.O., Timmerman, M.A. and de Jong, F.H. (1990) Mol. Cell. Endocrinol. 74, 125- 132.

Huylebroeck, D., van Nimmen, K., Waheed, A., von Figura, K., Marmenout, A., Fransen, L., de Waele, P., Jaspar, J.-M., Franchimont, P., Stunnenberg, H. and van Heuver- swijn, H. (1990) Mol. Endocrinol. 4, 1153-1165.

Klaij, LA., Toebosch, A.M.W., Themmen, A.P.N., Shimasaki, S., de Jong. F.H. and Grootegoed, J.A. (1990) Mol. Cell. Endocrinol. 68, 45-52.

Knight, P.G., Beard, A.J., Wrathall, J.H.M. and Castillo, R.J. (1989) J. Mol. Endocrinol. 2, 189-200.

Laemmli, U.K. (1970) Nature 227, 680-685. Lee, W., Mason, A.J., Schwa& R., Szonyi, E. and Mather, J.P.

(1989) Science 243, 396-398. Ling, N., Ying. S.-Y., Ueno, N., Shimasaki, S., Esch, F.,

Hotta, M. and Guillemin, R. (1986) Nature 321, 779-782. Maddocks, S. and Sharpe, R. (1989) Mol. Cell. Endocrinol.

67, 113-118. Maddocks, S. and Sharpe. R. (1990) Endocrinology 126,

1541-1550. McLachlan, R.I., Matsumoto, A.M., Burger, H.G., de Kretser,

D.M. and Bremner, W.J. (1988) J. Clin. Invest. 82,880-884. Merchenthaler, I., Culler, M.D., Petrusz, P. and Negro-Vilar,

A. (1987) Mol. Cell. Endocrinol. 54, 239-243. Meunier, H., Rivier, C., Evans, R.M. and Vale, W. (1988a)

Proc. Natl. Acad. Sci. USA 85, 247-251. Meunier, H., Cajander, S.B., Roberts, V.J., Rivier, C.,

Sawchenko, P.E., Hsueh. A.J.W. and Vale, W. (1988b) Mol. Endocrinol. 2, 1352-1363.

Oonk, R.B., Grootegoed, J.A. and van der Molen, H.J. (1985) Mol. Cell. Endocrinol. 42, 39-48.

Risbridger, G.P., Clement& J., Robertson, D.M., Drummond, E., Muir, J., Burger, H.G. and de Kretser, D.M. (1989) Mol. Cell. Endocrinol. 66, 119-122.

Risbridger, G.P., Sun, X.R., Klein, R., Hedger, M.P., Robert- son, D.M. and de Kretser, D.M. (1991) North American Testis Workshop, MontrCal, McGill University, Montrial, abstract 68.

Rivier, C., Cajander, S., Vaughan, J., Hsueh, A.J.W. and Vale, W. (1988) Endocrinology 123, 120-126.

Roberts, V., Meunier, H., Sawchenko, P.E. and Vale, W. (1989) Endocrinology 125,2350-2359.

Robertson, D.M., Hayward, S., Irby, D., Jacobsen, J., Clarke. L., McLachlan, R.I. and de Kretser, D.M. (1988) Mol. Cell. Endocrinol. 58, 1-8.

Robertson, D.M., Giacometti, M., Foulds, L.M., Lahnstein, J., Goss, N.H., Hearn, M.T.W. and de Kretser, D.M. (1989) Endocrinology 125, 2141-2149.

115

Rommerts, F.F.G., Molenaar, R. and van der Molen, H.J.

(1985) Methods Enzymol. 109, 275-288.

Saito, S., Roche, P.C., McCormick, D.J. and Ryan, R.J. (1989)

Endocrinology 125, 898-905.

Shaha, C., Morris, P.L., Chen, C.-L.C., Vale, W. and Bardin,

C.W. (1989) Endocrinology 125, 1941-1950.

Sharpe, R.M., Kerr, J.B. and Maddocks, S. (1988) Mol. Cell.

Endocrinol. 60, 243-247.

Smith, J.C., Price, B.M.J., van Nimmen, K. and Huylebroeck,

D. (1990) Nature 345, 729-731.

Sugino, K., Nakamura, T., Takio, K., Titani, K., Miyamoto,

K., Hasegawa, Y., Igarashi, M. and Sugino, H. (1989)

Biochem. Biophys. Res. Commun. 159, 1323-1329.

Symes, K. and Smith, J.C. (1987) Development 101, 339-349.

Teerds, K.J., de Rooij, D.G., de Jong, F.H. and Rommerts,

F.F.G. (1991) J. Androl. 12, 171-179.

Toebosch, A.M.W., Robertson, D.M., Klaij, I.A., de Jong,

F.H. and Grootegoed, J.A. (1989) J. Endocrinol. 122,

757-762.

Vale, W., Rivier, J., Vaughan, J., McClintock, R., Corrigan,

A., Woo, W., Karr, D. and Spiess, J. (19861 Nature 321,

776-779.

van Dijk, S., Steenbergen, J., Gielen, J.Th. and de Jong, F.H.

(1986) J. Endocrinol. 111, 255-261.

van Laar, J.H., Voorhorst-Ogink. M.M., Zegers, N.D.,

Boersma, W.J.A., Claasen, E., van der Korput, J.A.G.M.,

Ruizeveld de Winter, J.A.. van der Kwast, Th.H., Mulder,

E., Trapman, J. and Brinkmann, A.O. (1989) Mol. Cell.

Endocrinol. 67, 29-38.

Ying, S.-Y. (1988) Endocr. Rev. 9, 2677293.

Documents

Testicular Leydig cells in vitro secrete only inhibin α-subunits, whereas Leydig cell tumors can secrete bioactive inhibin